Table 3

Independent determinants of the prevalence of current injecting and the incidence of injecting transitions, using 13 620 visits of 996 drug users in the Amsterdam Cohort Study

	Current injecting OR	Initiation OR	Cessation OR	Relapse OR
Year^3-151	1.17	0.57^3-150	0.71	0.80^3-150
Year^3-150Year^3-151	0.96^3-165	—	1.05^3-160	—
Female	0.64^3-165	—	—	—
Non-West European ethnicity	0.23^3-165	0.32^3-150	1.60^3-160	—
Non-Dutch nationality	1.95^3-165	—	—	—
Age^3-152	—	0.61^3-150	—	0.85^3-150
Duration of injecting^3-152	NA	NA	—	1.14^3-150
HIV positive	1.40^3-160	—	—	—
Duration stopped injecting^3-152	NA	NA	NA	0.81^3-165

Current behaviours: ^3-153
Non-injecting heroin use^3-154
<daily	0.97	3.48	1.34^3-165	—
daily	0.80^3-160	7.13^3-160	1.68^3-165	—
Non-injecting cocaine use^3-154
<daily	0.81^3-165	—	—	—
daily	0.91	—	—	—
Tranquiliser use^3-154
<daily	1.33^3-165	2.44	—	1.40^3-150
daily	1.22^3-160	3.47^3-150	—	0.98
Prostitution	1.63^3-165	—	—	—
Steady sexual partner
none	1	1	1	1
currently injecting	1.56^3-165	3.35^3-150	0.70^3-160	1.17
stopped injecting	0.58^3-165	2.40	1.22	0.87
never injected	0.66^3-165	0.89	1.22^3-150	0.67^3-165
Daily methadone	0.76^3-165	—	—	—
AIDS diagnosis	—	—	1.80^3-160	—
Needles via exchange
none	NA	NA	1	NA
1–99%			0.67^3-160
100%			0.50^3-165
Frequency injecting
<daily	NA	NA	1	NA
daily			0.47^3-165
>daily			0.38^3-165

↵3-150 , p<0.05;
↵3-160 , p<0.01;
↵3-165 , p<0.001; GEE was used for significance testing, using a logistic link function; calendar year was forced into the models; NA, not applicable.
↵3-151 OR expressed per five year increase in calendar year.
↵3-152 OR expressed per category of years (see methods).
↵3-153 For injecting transitions, current behaviours were measured at the last cohort visit preceding the transition. For prevalence of current injecting, they were measured at the same visit as current injecting.
↵3-154 Reference group is no use of drug.